We consider technique for parallelization of computation with hypercomplex numbers in multidimensional space for signal processing purposes. In particular, parallel algorithms of the multidimensional hypercomplex discrete Fourier transform are developed. Results of their experimental research are shown.
In this paper, we offer and discuss three efficient structural solutions for the hardware-oriented implementation of discrete quaternion Fourier transform basic operations with reduced implementation complexities. The first solution-a scheme for calculating sq product, the second solution – a scheme for calculating qt product, and the third solution – a… (More)
The methods of fast parallel calculation of multidimen-sional hypercomplex discrete Fourier transform (HDFT) are discussed. The theoretical and experimental estimation of parallelization efficiency is given. It is shown that proposed method has very high efficiency (up to 90%).